Glaser, Strauss, Charmaz, Nelson, Claude.ai? When digital nomads use generative AI to build...

[music] Hello, Nick, my man. How you doing?

Recker, my brother. Hey, you're looking pretty uh intelligent in your glasses there.

Yeah, I uh I'm at that age now where I moved from zero glasses to three. I now

have three different glasses and three different rooms. They all have different strength and uh I don't want to live without them anymore. I mean like this is the world is open to my eyes and it's it's a much sharper experience than what

I had before.

Yeah, that's funny. So I have reading glasses so it's closed. I do that and then I have glasses for far but the computer screen is perfect. I don't need glasses for the computer screen. So uh

during the day all day I don't wear glasses but then if it's far near No, this is a special glasses for that sort of distance between me and the screen kind of thing which is which is amazing. [laughter]

amazing. [laughter] But yeah, so the story is of course I went to the I went to the eye doctor and said like, "Look, I I feel like my my my my uh my vision is getting worse." And

she's like, "Okay, let's test you." And

then she comes out with the test results. Look, Mr. Recker, your eyes are

results. Look, Mr. Recker, your eyes are 100%.

And I'm like, "But I don't see anything anymore." [laughter] She's like, "Yeah,

anymore." [laughter] She's like, "Yeah, that's just that's just what it is now at your age."

And then you just glasses.

You know what? This is the first time in your life you looked like a professor, right? Usually you look like You look

right? Usually you look like You look like kind of a dumb jock, right? Like

someone who should be with a basketball.

And and if you would only get a little gray in your beard, if you would stop dying your beard and you would just let the gray come out, then I think you'd actually look professorial at this point.

Well, speaking of people that look professorial and don't wear glasses, um we have one of them on on today. We have

our good old friend Daniel. Daniel, you

we know each other for 20 years maybe.

Feels like a long time. Anyway, Daniel

has a little bit of, you know, sophisticated bit grainous beard. But do

you have glasses? You don't have glasses?

No. Surprisingly, I I I don't know your age, Nick, but I know that I'm older than Yan. And all of my friends,

than Yan. And all of my friends, everybody in my family has glasses. I I

never had glasses. Not sure what's going on. Yes. And I I I stopped fighting

on. Yes. And I I I stopped fighting against the gray in my beard. So, I have a gray beard lot. And I I have to say, Yan, you do look good with the glasses.

You look a little bit like the smart emoji, right? The one with the the nerd

emoji, right? The one with the the nerd glasses. So, I'm not sure whether this

glasses. So, I'm not sure whether this is a good thing, but yeah. Uh looks what So, you're older than Recker.

You You look younger than Recker. Well,

yeah, you look younger than Recker. It

looks like you have smoother skin and and probably more hair. So, yeah.

Thanks. So, I can give it away. My next

birthday is going to be 50. So, uh that is in certain age.

So Daniel, maybe maybe the the reason that you look so healthy and young and everything is that you have a enormously awesome uh lifestyle. You are in fact not only a digital nomad, right? And you

you whenever we catch you, you're at different places on this planet in different different continent, different countries anyway, usually in the same region. But uh I find it awesome like

region. But uh I find it awesome like you are a digital nomad and of course you've you're studying digital moments.

You've written a couple of papers on them as well. Um but how is being an IS professor as a nomad?

Uh well, it's an it's an aspirational it's an aspirational goal. So I I aspired to the to the uh um nomadic professor uh as an as an employee there's some limitations how much

nomadic you can be but because I have been studying this for the last 10 years that also means I and I do ethnography which means you know you live with your phenomenon you travel your phenomenon.

So I try to be in nomad land as much as I can and uh I'm in Daang Vietnam at the moment looking at the co-working spaces

um and things here right now.

Who is the world's most famous nomadic professor in throughout history?

Oh throughout history. That's a good one. Uh

one. Uh I think it's Paul Adidos. So you know Adidush. E R D O S. The Adidos number. a

Adidush. E R D O S. The Adidos number. a

mathematician machine graph graph networks. So yeah, he was uh I

graph networks. So yeah, he was uh I guess he was a heroin addict. He was a mathematician and he would just live on other academics couches and stuff. He

would be here at Notre Dame a lot actually and uh he thought we were a very mathematical university because everywhere there was the plus sign, you know, the cross were a Catholic

university. Uh, and yeah, he would solve

university. Uh, and yeah, he would solve people's math problems and live on their uh, and like he was totally nomadic. You

don't know this guy? [laughter]

I'm not aware in particular, not of the the nomad story, but okay, that's interesting. So,

we'll we'll solve maths for a shot, right? So yeah, that didn't he also go

right? So yeah, that didn't he also go to Southeast Asia because that was one of the things that established the the hippie trail which is kind of the original gangster of uh independent traveling and nomadism where a lot of

the you know the hippie trail from London and Berlin in the 70s kind of following the Beatles uh uh to India and then all the way to you know Kathmandu or or Bangkok and a lot of them were

actually in search of cheap heroine back in the days. Yeah. So that was connected to the uh emergence of digital nomadism you could say. Even of course it wasn't

called like that. It was called hippies uh vagabons or later backpackers. Right.

And it's all the baby boomers. So these

people became professors later on.

Right. Uh probably.

I I do believe that uh I I I hope I got that right. But I think Shirley Greger

that right. But I think Shirley Greger actually did the the hippie trail. Yeah.

So that was some is professor that we all know.

But the So when I think of nomadism, I think of the beach, right? The the the the movie or the book. Uh is is that is that still apt as a as a metaphor for

your your life and your work?

Ah, for me personally, um you know, I'm I'm I think [clears throat] you're Australian too, right? I'm I'm I'm Aussie now. So we kind of um bit spoiled

Aussie now. So we kind of um bit spoiled with speeches. So um not [clears throat]

with speeches. So um not [clears throat] for me personally and for the nomads in general I think you know this is nomads actually typically don't like to be you know it's a very individual

individualist lifestyle right so they don't like a collective label because that's precisely not what they want to be right if you want to be part of a community you stay in your community you don't be the one in 100 that you know

lives their individual I go to India I go to whatever um so they don't typically like it and it's a very kind of mixed group of people. It's a little

bit like don't call me a hipster, right?

Even you live in, you know, Batford Avenue and have a beard and I'll barista. No. No. But you're not a

barista. No. No. But you're not a hipster. No.

hipster. No.

So, wait, are you are you married? Are

you married and do you have children?

Uh, I have a child. Yes. and I'm editor of a book called uh word schooling which is homeschooling but for the less religious and more nomadic among us where you essentially use the cultural

experience of growing up around the world as part of the educational journey and then of course you need to see you know what kind of degrees do you get for the little ones is it going to be an IB

and uh you know so there are actually solutions for nomadic families and since uh we actually travel on three passports

I'm German. My baby is um Australian. My

I'm German. My baby is um Australian. My

partner is Filipina. So uh we travel on three different passports. So I I do think to to the limited degree that an academic career allows it, we are

somewhat in that nomadic space. Yeah.

And um yeah, with the beach, I think that's that's of course the the AI cliche image, you know, from the LLM about digital nomads and you know, laptop at the beach uh comes up.

Yeah. But in many cases actually digital nomads like places that are away uh from the beach like Chiang Mai Island uh I wrote a case about that is I would say the most popular destination median

which I have also been quite often is also not really at all at the coast because I guess it's the same like living in Sydney. How often do you go to Bondi? Yeah,

Bondi? Yeah, twice a year. You know, it's, you know, it's good for a holiday, but if you're actually working, you're looking

probably more for moderate climate, safe environment, good air quality. Um, so

it's not as beachy as the the image uh might be. Also,

might be. Also, why bother being do you, you know, why bother being a nomad? If

you want a safe air quality, all that, you could be in a suburb in in whatever major city in the world. And so what's the advantage of the being in Chiang Mai as opposed to in Sydney in a suburb if

you have your kid?

I think one is the search for the other, right? The search for the orient. Um so

right? The search for the orient. Um so

you want to be in environment that's [clears throat] not how you how you grew up. So I think I I share this and I know

up. So I think I I share this and I know that a lot of digital nomads when I interview them they're always looking at me, you know this, you you're one of us, right? Why do I need to explain this? I

right? Why do I need to explain this? I

I say explain it for the rest of us, right? So why would you make your life

right? So why would you make your life complicated? um going around when you

complicated? um going around when you have a home community with family and there can be different reasons. So

typically it's this um push and pull or um who was this Frankfurt school um uh I think fre uh with this freedom from and

freedom towards right it's freedom from the the cut out lifestyle you know living like my parents did suburbs buying a house blah blah blah merits

corporate cubicle trajectory career stages and then freedom towards is essentially the unexpected falling a laugh on the other side of the planet, cultural, experience, you know,

optimizing against experience instead of um, you know, money for example in life.

I think that's something that's that's widely shared and that's why I'm typically moderately positive on digital nomads possibly because I somewhat

identify that, but also because a I I think there's a legitimate case for this to be made which is kind of postmaterial which I I appreciate. uh as well as that

I'm having lived long time in Thailand I look at it also very strongly from the receiving countries perspective and for them this is a straightforward transfer of economic um activity it had a lot of

negative effects but of course overall the key aspect is suddenly you have e economic activity you have all the people that you know run co-working spaces tech startups they go co-learning

events every week you would just not have that uh if the digital nomads wouldn't be coming that's Why typically people in the west don't really like digital nomads because they say well you know they're taking advantage of our

system and of their system so they're kind of freewriting the global system but from the perspective of the local communities it's typically more welcomed that's why now I think we are now I have

a database I think it's 51 different digital nomad visa available including Thailand we have three uh uh for digital nomads one I was directly involved in uh

uh creating it And the hope is to get more digital nomads in. And we spoke about that because you want to get young people, you want to get, you know, the spending and the spending is long and

wide. They're not on package tours when

wide. They're not on package tours when COVID came. They were not back in New

COVID came. They were not back in New York or back in Shanghai in two weeks.

They s, you know, they were staying in Bali or uh in Chiang Mai keeping the lights on uh to some degree. And you

know, they're in local Airbnbs, local villas, spending in restaurants. There

is economic inequality still but probably not more than with normal kind of visitor economy tourism and of course they stay longer and have a somewhat

higher investment. That's at least why I

higher investment. That's at least why I think many countries see that as a as a positive. And now that Korea and Japan,

positive. And now that Korea and Japan, particular Japan, is a bit surprising because they're not necessarily known as a number one immigration country on the planet, but they have a

very strong digital nomad initiative.

And I think that's probably, you know, falling fertility rates. If you don't have your own babies, then, you know, getting somebody that's probably university degree and comes with business and tech knowledge into the

country is, you know, the the least worst uh uh alternative.

Daniel, you totally you totally got me hooked on digital nomadism. I have to admit I had no idea. I really know. So,

so maybe just answer this uh cuz you want to get to a different topic, but if I wanted to start reading up on digital nomadism, what's a what's a great paper to start? It could be one of yours,

to start? It could be one of yours, could be, I don't know, one that you like. Give it give us a shout out to a

like. Give it give us a shout out to a really good paper that explains what it's about.

I see that you have multiple.

Yes. Yes. I I do believe that I actually was the first to use the term digital nomadism and I reserved the domain

digitalism.com. digital nomad as a as a

digitalism.com. digital nomad as a as a noun is from um I think 1997 uh uh uh book but digital nomadism I thought is better because you know digital nomads

then makes it centric on the digital nomads themselves rather than the total phenomenon that's why you know the anthropologists always prefer orientalism tourism you know it's isms

so I felt that's a better title um good papers okay so there's definitely some from my own group that I want to shout

out which is uh I think Blair's uh and my paper on emancipation or the the version of emancipation and self-actualization self-realization that

digital nomads u enact. So kind of why are they actually doing that?

Essentially key thing is that they're not [clears throat] looking for solution to a more meaningful life by you know having the red flag and waiting for social revolution that never comes

and you know look looking at it as a class problem and collective action problem but they say okay let's game that system and reverse engineer it and optimize it in a more single

entrepreneurial. So definitely

entrepreneurial. So definitely individualist solution instead of a collectivist solution but to many of the same aims. So you want to be financially independent, you want to have meaningful

work, uh you want to have time for other things um than work. So it's not entirely a completely different thing.

But we had to redefine what emancipation means because of course that comes out of critical theory and critical theory was in a time where the reference framework was the nation state. But for

digital nomads the reference framework is no longer the nation state right if Frankfurt school marks wrote about Germany or France right they did not consider the option

that you could just essentially remove yourself from from statehood and become kind of citizen of no really that they have like five passports and you know

where to which country do you belong to none really yeah so it's a it's a bit like the the Bitcoin stuff, it's kind of in in something that I like to call post Westfailian, like something that's in a

socioeconomic space that's post the West Failian system, post the nation state, right? I I failed to put better

right? I I failed to put better terminology to it, but it's interesting because it's almost like a foreshadowing, but what comes when you know Germany is only like what

150 170 years old as a nation state. So

the nation state concept wasn't there forever, right? And maybe if you're

forever, right? And maybe if you're building increasingly stuff based on an internet infrastructure, maybe the nation state concept.

I don't know if the nation state was not there forever. The nation state has been

there forever. The nation state has been there for quite some time. And nomadic

lifestyles have been there for quite some time as well. I mean, we've had Roma, right? The gypsies. We've had uh

Roma, right? The gypsies. We've had uh here in the United States, you know, the Native Americans. I think there were a

Native Americans. I think there were a lot.

We had them in Africa for thousands of years. Yeah. nomadism was pretty

years. Yeah. nomadism was pretty I think that uh it was about last year that the number of digital nomads is now higher than the number of traditional nomads all combined on the planet which

I thought is interesting and uh also I find interesting that they're still clashing with everybody else almost similar like the Jewish diaspora which kind of always you know is in that same

globally relatively uniform but the very niche minority in each society that they with the exception of Israel obviously but um somewhat related problems of you

know typically somewhat higher performing and so forth which then also create sometimes frictions. Uh so that's the closest equivalent the traditional nomads and the Jewish diaspora that I

could come up with. But of course it's also very strongly produced by the internet. Right. So there's I don't

internet. Right. So there's I don't believe in technological determinism but in technological agency right you put the internet onto an island with a new society and then eventually you will

have social media influencers and digital nomads be because the internet wants to be used that way. So in a certain sense it's almost a necessity of of the invention of the internet to buy

our goods certain that's a I'm sorry but I really no I'm I'm still hooked by this. You're saying

that there is technological agency in the sense that something like social media is inevitable through the diffusion of the internet because the internet wants to be used that way.

That's an interesting statement.

I would I would say let's let's put it like that as a thought experiment. You

drop social media onto a society that has not social media before. You let

them run or you let 20 such societies run. You come back after 10 years, I

run. You come back after 10 years, I guarantee you there will be social influences. So I don't think it's really

influences. So I don't think it's really the social influences using the technology. It's really the technology

technology. It's really the technology wants to be used in that way and produces the social phenomena like the I I think that's close to what we call you know reversal of the

ontology, right? So anyway uh yes. So

ontology, right? So anyway uh yes. So

that's that's how I look at it.

[music] All right. So let's shift gears from

All right. So let's shift gears from nomadism a bit and uh Recker, you were telling me about something you sat in with Daniel with the SIG GTM and uh

where they were talking about grounded theory methodology, right? Which is of course a qualitative methodology for typically qualitative. There are

typically qualitative. There are quantitative versions of it but typically qualitative methodology that uh you know you you start with the data you start with the interviews the

ethnography whatever and you elicit you do uh comparison with the literature you do rounds of coding and you elicit new insight you generate new theory and you

know of course so what's the role of of uh AI in this space right yeah I mean I mean we we've uh Danny and I got talking around must be almost half

a year ago just around the the last ISIS before during and after and the reason is that that Daniel is not only a researcher on digomatism but you're also the chair of sik GTM which is I'm not

sure whether it's the oldest but it's one of the oldest it's certainly one of the largest and certainly one of the most popular um special interest group of the AIS um GTM and we've been

chairing the the annual workshops which traditionally run on the last day at ISIS if I if I'm if I'm not mistaken. Um

and this past year um I didn't attend because I I had to fly home early but you ran a panel on GTM research in the age of AI and because this this is an

obvious question right I mean we've we've seen um AI taken over various parts of science. Um but I find it fascinating to think about it in the

context of what what is to me a very manual, very humanistic, very human, very interpretive, you know, whatever type of very very person focused uh

discipline, right? So if you think about

discipline, right? So if you think about where GTM comes from, um it's kind of like a human a human method. Um yeah so uh like you talked about ethnography

where you have that literal eye so you subjective immersion is very important uh uh much of that in in in in inductive data analysis and you being there and you know you studying things and you

observing things and you taking field notes and all of this becomes data that you later code and develop theory from.

I mean, all of that seems to be a cornerstone of um GTM, except that now I can just have have a tool do it for me.

I mean, take us take us into this panel.

I wasn't there. I believe Nick wasn't there either, but you know, that must have been a fascinating discussion.

Yeah. Uh I have to say uh I love Nick's directness because every discussion at the moment somehow ends up in the end about AI. Anyway, we we tried to do

about AI. Anyway, we we tried to do something about remote work. I was doing a a course with Microsoft and it was supposed to be remote work and hybrid work, but as soon after 15 minutes, somebody dropped but what about AI and

then the rest of the discussion was so I I love Nick, you know, let's just skip the foreplay. Let's go straight to AI.

the foreplay. Let's go straight to AI.

[laughter] So, um yeah, so yes, thanks Jan for and Nick for the introduction.

I'm actually here as as president of Sik GTM and wants to tell you uh and uh thanks for the interest a bit about uh Sik GTM and you gave essentially like a

elevator pitch summary of what Sik GTM um is uh Yan and yes it is uh I would say a very humanist uh and humanity side

so in in my mind I don't really talk about interpretivism or positivism I'm more talking about are we seeing our kind of intellectual heritage as coming

from science, capital S science, philosophy of science and the natural sciences or do we see it coming as from the humanities uh anthropology to

sociology side which gives us a completely different um history uh and with uh AI of course the question is how

does that fit? Yeah. because there is a certain I mean a it's absolutely impressive. So, Supra gave uh a keynote

impressive. So, Supra gave uh a keynote on you know running through how you can essentially use LLM currently and I I

test wise do that even I do still have some misgivings including that of course as a very qualitative you know I'm almost falling off as an adnographer on

the left side of the spectrum right so I very strongly believe that the text and the author need to be looked at as one consistent thing so It's not that hypothetically

the knowledge or the experience or whatever is described in a text could be known by the author but maybe they don't only the AI produced it and maybe it's not wrong but in a certain sense I

expect the author of the text to know that and to have found that out and consider that to be the best interpretation of the situation. That

being said, yes, there's definitely um much that can be done uh with LLMs in particular the the step that we have um

typically performed manual. So I guess there's two ways of looking at it. Did

LLMs just essentially kill qualitative inquiry? particular anything that's kind

inquiry? particular anything that's kind of data and coding uh based and essentially where AI can take a year of

work and and do it in in seconds or did we just essentially in invent SPSS or regression tools for qualitative analysis and now we're finally on par

with the quant guys which had this massive tool support and which say oh I had you know uh 5 million data points no problem Right. I never have 5 million

problem Right. I never have 5 million data points. Yeah. Because I talk to

data points. Yeah. Because I talk to people, I fly there, right? I sit in a co-working space and I have one case study after 9 months. So that's not very

efficient if you're looking at a KPI metrified academia because it's very slow. So maybe alm gives us a boost,

slow. So maybe alm gives us a boost, right?

So So you did this panel and people were talking about it. Uh and I guess it stands to reason that we're we're going because it's about language, right?

Right? An LLM is necessarily a language model. And when we do grounded theory,

model. And when we do grounded theory, we're essentially coding language. So

that's kind of what LLMs do. So what is the I mean when you guys saw this panel, what Yeah, clearly it can be used.

Clearly there are some challenges to using it for GTM. Were there anything that jumped out at the two of you specifically from the panel that was like, "Okay, these are the things that we should uh that it's good for and these are the things that maybe we need

to be cautious about."

Wait, Daniel, let me rephrase that question. Was there anyone seriously at

question. Was there anyone seriously at the panel saying we should ban it? Was

there anyone having a clear antagonist anti-I stance at the panel saying like GTM shall never be done with the help of AI?

Was that even a position at all?

Uh no I don't think anybody uh uh said that. I think what and with my my guest

that. I think what and with my my guest third head on editor of of JIT, I think what I definitely always would want as a

GTM reviewer or as an editor is transparency about exactly how it has been used, right? because one is possibly a very efficient very clever to

be debated but I I think in general completely fine approach if you can justify for which ends you used it uh the other is academic cheating right if you say look I wrote this and I

performed the analysis and actually the AI that you know that's straightforward academic fraud so that's why I think nobody is against AI use I think I would

be and everybody would be against undeclared or unjustified particularly undeclared essentially pretending that this was human analysis when it was not.

So I think transparency in in terms of what was done by whom or by what um needs to be there and then I think there's also um and that's also

something Yan and I discussed and we had this idea too that I might come on the podcast in terms of that typically people think of already qualitative

research as more or less one thing right that's like I would say this is like Asia being one thing right you know you your Arabs and Japanese are are really

just kind of the same, right? No, really

not. It's it's really like very very different. And even within GTM, there

different. And even within GTM, there are very different camps, right? So I

would so most people essentially uh and I guess if I can kind of force one point into this podcast is typically GTM researchers are uh the worst nightmare

of other GTM researchers because the terminology GTM suggests that it's one uniform thing and then everybody applies the quality criteria of their particular subtype and you know the same happens

with this qualitative research in general but I would say there's definitely the the Glazer and Strauss difference. So the original Glazer and

difference. So the original Glazer and Strauss and particular Glazer later was very inductive focused right. So

essentially starting from the data from the phenomenon but Strauss has a much more kind of uh you know um I would say can in a sense like pragmatic theory

informed processdriven uh approach abductive reasoning you could say and I'm actually constructivist GTM researcher so I use

this very light I essentially use this as a kind of index to my data but I'm taking a very interpretativeist constructivist approach more like

ethnography or you know like qualitative research as essentially focus on on interpretation rather than uh kind of rigid the evidence forced us to theorize

it in that way uh approach and then of course you have the arguably forced journal which people sometimes call computational GTM or uh you know the

reconstruction and sometimes people try to avoid the term GTM not the least because many of those are actually coming more from a big data computer science and other roots not actually

from a qualitative upbringing but for all intents and purposes it's really the GTM approach so that means we have already four different approaches and

the acceptability of the AI step is actually very much depending on the particular subtype that you're subscribing to so if you're doing pure

induction and I would say so the clea type and the computational approach then I think it's more acceptable because you just want to identify interesting

patterns in huge chunks of data and being able to process more data um which I think AI is really what it can help us

with uh is straightforward um a positive of course if you're more like an anthropologist and want to precisely understand are actually the ties and the

Indonesians and the Colombians and the digital nomads and the Colombia University, Connecticut, New York group which are talking about the nomads are

actually kind of aligned or do you need to consider their cultural histories and supports to really make sense of their interpretations. This is really

interpretations. This is really something where AI still really fails because it just doesn't have this kind of what we call in German meaning

understanding. I have still not seen

understanding. I have still not seen good results in that it typically glosses over stuff and also tries to is always a bit regressive. It tries to go

back to the logics that we have. It's

typically in its higher level theoretical analysis not very good in saying oh what this really shows what we believed so far is wrong and we need to look at that differently. It always

tries to go back to the training data and force it into no no that's that's the theories that we believe in which again uh is for some approaches like a

more you know scientific or deductive approach probably if you have this mental model of you know continuous improvement you know uh building on each

other standing on the shoulder of giants that's fine but if you're doing the more kind of problematizing um the kind of breaking the paradigm uh

or just interpreting it from a completely different framework said okay this is all ultimately western uh viewpoints let's cut this and let's take

uh you know how about Buddhist ontology why not so essentially for this kind of things it's really bad so far from all I have seen but I I guess I'm pragmatist

enough with you Nick to say well the proof is in the pudding if you show me analysis where this is well done and where it provides something which is meaningfully contributes to GTM and I

think that will definitely happen at the coding level. I'm not so sure it will

coding level. I'm not so sure it will happen at the theory construction choice and you know like the the the 20% that I

I think still needs us in a certain sense. So Daniel is there at at GTM I

sense. So Daniel is there at at GTM I mean from or at JIT have you seen work that explicitly and transparently really

says like here is basically an almost fully or to the largest extent possible automated GTM version where I collected some data and I chucked it on into a tool you know and then maybe I wrote

about it maybe I had AI help me write it whatever but where the main part of the ground theory methodology was in fact automated is there like that. I' I'd

like to see it. Like I haven't seen it, but I I probably don't come across so many as as you do.

Automated with an LLM is the question, right? Because I think you know there is

right? Because I think you know there is quantitative GTM so you could argue some of that's automated. Uh but

uh how about yeah an LLM used to analyze data. Have

we seen that with qualitative data? Have

you seen that in JIT?

Uh not in papers that I have personally seen. Even for example, Supra showed a

seen. Even for example, Supra showed a hypothetical example, but I haven't really had one paper under review in in my hands that that fully committed to

that approach. I think Max QDA, which is

that approach. I think Max QDA, which is the German version of in in Australia, we use Envivo and you know it seems to be depending slightly on in which institutional environment you are, but I

think Max QDI now introduces [clears throat] feature.

Yes, it does.

Yes. And I think that and Bivo had like auto coding for the longest of time but that was still kind of the old approach not the LLM approach but more like semantic mapping. It didn't really work

semantic mapping. It didn't really work for me. Yeah. So I haven't seen a paper

for me. Yeah. So I haven't seen a paper like that but I think you know there must be coming any they must be out there. So I I wanna I want to have I want to make two observations and see if you agree with

them. So number one on this particular

them. So number one on this particular point is I think they're out there. I

think from what I see is people are actually trying to hide their LLM use in papers. Not specifically GTM certainly

papers. Not specifically GTM certainly in GTM. I I would wager even more so in

in GTM. I I would wager even more so in qualitative or inductive GTM type of work. My uh my observation is that

work. My uh my observation is that probably happens in all papers that that people downplay, minimize or literally hide away the use of AI if at all. I

have not seen really an forthcoming outright very explicit we used AI in the research process we couldn't have done it without it sort of thing. So that's

observation number one. See whether you agree with this. And the second observation or more like a prediction is is based on your point that it's sort of like a a tool like the introduction of

you know smart pls and lis and all the SPSS were for the quantitative researchers and that we now have uh you know more capable computational software for this type of research process and

what that has done to quant is something that I would expect to see there. namely

that now this will mean that over time the requirements of for such studies especially in terms of quantity quality enrichment of data will go up yeah as you have better tools and as more

analytics become possible the natural consequence would be well then I'm upping the ant for what does it mean to have a good study that I don't know 20 interviews are not good enough it'll be

100 now um I want to see you go through 20,000 pages of documents because you you can or you can so much more easily than before. So, I thinking we we're

than before. So, I thinking we we're upping the the study requirements because we have better software support.

Yes, I think that's a fair way of of um looking it. it it creates a little bit

looking it. it it creates a little bit uh uh and that's not specifically only uh GTM that that's about LLM used in many genres of papers including of

course literature review conceptual papers which are all now kind of as a very genre problematic for one reason that it takes quite a lot of time to

distinguish a really good paper from an almost good paper that will never be any better because it was ultimately just AI supported and generated. And essentially

um now it becomes essentially so cheap to make a potential conceptual paper or potential literature review paper. You

can just generate them but it still takes the full amount to go through all of the hoops to do the quality check at the end which is a big burden. I can

tell you that as a journal editor and probably you do editing right? So you

would know that now essentially the the onus of doing the work and making sure everything is proper was supposed to be with the authors. Now essentially yes

the authors can kind of cheat, right?

And then the quality control shifts to the reviewers and editors which kind of is is the the paradox, right?

becomes much easier to produce but it still needs the same amount for each individual item to be revealed or be probed as which is problematic. Not so much for each

problematic. Not so much for each individual researcher or piece, but at the systemic level. It's kind of the double submission of the AI age, right?

I'm a bit worried about the sustainability of the overall system. or

Natalyia Levina is of the view that conceptual and literature review papers will essentially go out of business because unfortunately they can be too easily faked and whether somebody puts

their heart and mind for years into a paper or somebody essentially wrote a clever script and has some rough ideas around it. uh I think the threshold in

around it. uh I think the threshold in quality you would need to dig really deep to see the quality difference between these two types now and given how everybody is now getting flooded

with these kind of papers it's really creating a problem but I I think to to to step one thing back and I had a conversation with uh sim le office the

other day and I think how he looks at the stage that we are at with LLMs and AI is a little bit like his own

invention of TCP IP and www 3 35 years ago, right?

It's very difficult to predict where this will go, right? It's it's you know, how would he have predicted Uber and LLMs and uh Tik Tok, right? So,

right? So, it's it's you know, even as president of Sik GTM, I cannot give you even for the GTM space a confident estimate where this will be

in 10 years. Yeah, I do think that your issue about reviewing, right, people can generate a lot more content and we're reviewing manually or at least we should be reviewing manually and that's a problem. It's not sustainable and I mean

problem. It's not sustainable and I mean the one answer is the market the reviewing marketplace which I think we have to do as a field at some point in our lives. uh we can't have three teams

our lives. uh we can't have three teams you know you send the paper to MISQ it gets rejected you send it to JIS it gets rejected you send it to JIT it gets reject that it's three different groups

of people who have now looked at the paper and rejected it right that's their time that they're that they're focused on so we do have to solve that the other thing and I just want to bring this up

Daniel just because I really like what you said that there's the Glazer and Strauss then you distinguished constructionist uh you I guess what you might call manual constructionist and

then there's the uh which you seem to align more with inductive which is the computational approach. So you had four

computational approach. So you had four approaches to GTM. Um and then later you kind of brought up pragmatism. I

actually think that G grounded theory people make so much nonsense about their particular glazarian and stressarian and do you start with codes? you have care.

Here's the deal. Eventually, you need to make a contribution.

Uh, right? Eventually, you need to say something. And that's something that you

something. And that's something that you have to say that you're going to say, you're going to make sense of your data.

Uh, that something that you have to say has to be with respect to what we already know. So, some assumptions

already know. So, some assumptions either from the literature or from conventional wisdom or something or existing theory, which is where we, you know, essentially codify our

assumptions. uh that we've learned from

assumptions. uh that we've learned from the past. So in the end you have to use

the past. So in the end you have to use your data and say something and your data could say many different things. So

your data doesn't just say one thing.

Your data you know it's the problem of underdetermination right from from a given data set you can do any number of different interpretations. So the point

different interpretations. So the point is there is no difference. we make such a big difference about we do you know everybody is so they're they're so partisan about this stuff and in

actuality in the end what do you have to do rigorously look at your data in order for you to question some existing assumptions extend that and say what we

know is not entirely the case here's how we can extend what we know that's a good contribution regardless of the genre uh right so to pretend as if these are

fundamentally different things is kind of uh I think unnecessary.

Uh I I would mostly agree with you uh as a fellow pragmatist. So I think in many cases and that is goes for all of is research. I think we not only have a

research. I think we not only have a theory fetish, we also have a method fetish where people get you know lost into the nitty-gritty did you call it chain of evidence or did you call it

this? Uh rather than looking at the

this? Uh rather than looking at the greater picture does that for example solve an actual problem in the world that we really have. Right? So I think

that we underemphasize the impact on practice right because we are we are essentially an applied field right it's not I go out with a grant proposal in

medical research and I say for $20 million I will c your cancer or for $20 million I will solve a highly theoretical problem that doesn't use is of any use in practical medicine

[clears throat] but I find it conceptually interesting no I will give the 20 million to the guys that solve cancer and I We have a as a field that was really created for a particular

practical space, we should much more place much more emphasis on the usefulness interestingness uh applicability of the knowledge that

we're generating than on the theoretical delta or the methodological rigidity.

And so far I agree with you Nick the point where I probably looking at a lot of GTM stuff of course see sometimes that we kind of and I guess it aligns

with your point to some degree. Uh I

think in an ideal world we should let everybody do their GTM approach as they see fit because there might be differences in you know do you allow

multiple interpretations multiple truths to exist or is there a single outcome you know there are some differences but let everybody do their favorite

approach. What practically happens is

approach. What practically happens is somebody does like I do constructivist GTM, right? Then somebody holds me to

GTM, right? Then somebody holds me to the much more rigid and scientific quote unquote standards of of laser and pure induction. And then the senior editor

induction. And then the senior editor goes, Daniel, don't be so insistent.

Just call it, you know, loosely grounded theory coding techniques. And I said, no, we should not be doing that. That's

kind of a regression because this is constructivist grounded theory, right?

So this is by all means what I'm doing.

Let me just call it. So I would say I agree with you that we shouldn't be so religious in enforcing our quality criteria on others. But then also let

people that are let let me assume I know what I'm doing.

Yeah. Let let me do it. Even your

particular favorite approach for GTM would be different. So I think it's a slightly different tweak on what you said, Nick. I'm mostly agree but I think

said, Nick. I'm mostly agree but I think it's not that we cannot say it's essentially all the same and you're just doing GTM. I think there are genre

doing GTM. I think there are genre differences but I think we would should let all of them stay. Same as films are films and I love films and I love all genre of films but you know you can't

apply the criteria of a horror film a good horror film to a good documentary because from a good documentary I just don't accept uh expect jump scares right

and from a horror movie I don't expect you know transparency and integrity in quoting right ex except that except that the difference between a horror and documentary team are not the the

difference between a constructivist GTM and Lazarian GTM, right? So these could be two different subtypes of well you know like I see the way that I look at it is like we are in a pluralistic field

and over not you know historically we have evolved to allow all sorts of different genres broadly speaking right so we had a I had a discussion in class yesterday where someone said like oh but

we don't allow ethnographies and I was like no here are five of them and then the person said no these are ethnograph but I'm talking autoethnographies as if that's a completely different thing and I said no that's actually not true here's a couple of these

So I think if we regress the genres to subg genres and sub subg genres and everyone has their own set of criteria that's also not very helpful. So let

there be 10 different ways of doing GTM but let there also be a standard set of criteria by which we we can say look this is a good GTM or not irrespective of whether you claim it to be that

particular kind or this other particular kind because otherwise anything goes. So

at some stage we need to regress on the subg genres. We need to stop regressing

subg genres. We need to stop regressing on the subg genres. story and agree on some more or less standardizable criteria. I mean you mentioned some of

criteria. I mean you mentioned some of them you mentioned look when we use AI it needs to be transparent it needs to be reproducible etc. Well that's holds for everything we do in science. So

that's not even a new criterion right we always it's called valid yeah it's called valid inference our whatever our contribution we're making to what we know already. So we have to

establish what we know already or else what are we doing talking to ourselves.

So in the end it has to be our contribution of what we know already and it should be based on valid inferences.

So that means logic that means if we're using statistics you know good statistical inference approaches if we're not using statistics good you know uh I guess logical and and rational you

know there are plenty of different uh approaches. All right but here's the

approaches. All right but here's the deal recker we got to roll. We have to wrap this up. Um, but I do want to ask Daniel one more question and this is

he's a editor over at uh JIT, Journal of Information Technology. And Daniel, I

Information Technology. And Daniel, I don't know if you've ever heard the Stefan Zidle uh theory on JIT. Have you

heard this?

Uh I I know Stefan Zidle series, but not the one on JIT. I would be very interested.

So I I'll probably get it wrong, but it's something like outside of our field, they don't know what information systems is. They think of us as

systems is. They think of us as information technology. And we have one

information technology. And we have one journal which is the journal of information technology. So if you're

information technology. So if you're outside of our field and this is one of the explanations for the great impact factor of JIT uh is that people outside

yeah it has the best name the it has the best journal name in our field and uh it gets a lot of halo effect and a lot of benefits because people will Google

search see JIT and think that must be the top IT journal. Uh let me quickly comment on all three points. So one uh Jan's point about with the horror movies

and documentary movies well the genre difference is a bit more than the sub terms in qualitative and it's a reducive argument right it's it's pointing taking the the point to the extreme to make the

point that even the genres yes the sub genres are similar but not the same in my view and I I would recommend Supra's paper on the different genres of qualitative research where you sometimes

see kind of review accidents where then somebody says you know this is an interpretivist constructivist story and then we looked at the coding and we calculated an intercoder reliability of

7.835 835 it doesn't go together right this is a kind of a jump scare in a documentary movie um that's what I meant and then um

Nick your point about uh JIT it can be a blessing and it can be a curse we also now getting last year I think we got 2,200 submissions 2,000

200 yeah our acceptance rate per our publishers comp uh calculation it's on the website it's 0.8% 8% because we are quarterly like MIS quarterly but we

getting more submissions and we only have four issues with about six to eight papers. So I guess there's a lot of

papers. So I guess there's a lot of people who hate us now because that thing but of course also we we do try to publish much more uh um papers. So what

I'm saying is yes the name is great.

Thanks Nick. I I I I think it really is.

On the other hand, we do get a lot of things that would be actually missubmitted in a sense because we're getting things that are too technical or you know because they're just seeing it

as it as a general label having no understanding of the is discipline and we do accept papers that might not be housed in the discipline but you would need to be generally in some sort of

sociotechnical space or or draw a link to that space otherwise it wouldn't be for our reader.

So yes, I think it's mostly positive.

Um, but it also comes with that we have to lot of sorting out to do. So that's

why I'm sharing my job as editor with with Wendy and Marco. It's it's we do get quite a lot of submissions. But I

need to ask one follow-up question on that. Is this one of these AI bumps that

that. Is this one of these AI bumps that we see in journal submissions? In other

words, has it skyrocketed over the last year or years specifically or has it gradually developed to such a high number? So for people outside like this

number? So for people outside like this is a very high number. This is higher than MIS quarterly as far as I know.

This is higher than JS. It's higher than most journals. In fact, uh uh we usually

most journals. In fact, uh uh we usually don't have journals getting 2,000 submissions as far as I know. Daniel

uh I'm not sure how much I think MIS quarterly does uh get uh north of a thousand at the moment, but uh yes. So

it it is it is quite uh high for that we are considering. So if MIS quarterly is

are considering. So if MIS quarterly is a big tent journal I think we are also a big tent but a small version of a big tent. So we are very you know not a

tent. So we are very you know not a purely qualitative not a smaller version of a big tent is a small tent small [laughter]

a small inclusive tent. Yes. Um yeah but but yes we we we do get uh the AI bump uh but it has also been going up. So I

think it's it's a dual combination of a growing field uh the name which attracts submissions that are from computer science or from

people that don't really understand uh you know our field uh and of course the AI bump the AI bump we therefore experiencing

kind of multiplied but I do think that all of the journals I've as the editors they're always talking so I I know what the other editors um Watsu or Supra or

uh Kieran or uh Robert are experiencing and everybody has the AI bump here.

That's that's a fact. And by for us this just means increasing a base of 7800 and to 2,000. For others that might be 400

to 2,000. For others that might be 400 to 800. Yeah.

to 800. Yeah.

Yeah. Wow.

It is a lot of papers. Yeah.

Thank you. I mean we we like there are three streams of conversation we could go on forever. The digital nomadism was fascinating. The GTM stuff is you know

fascinating. The GTM stuff is you know and the impact of AI is fascinating. JIT

and its role in the marketplace of ice journal is also fascinating. So I think we only touched upon all of them. So

maybe we'll have to bring you back on at some stage. [laughter]

some stage. [laughter] Yeah, there's always the there's always the risk if you're talking to me that I will be talking about digital nomadism.

Anyway, if I can put in the plaque. So

anybody interested in qualitative and GTM, we have this annual event on the last day at ICIS always the Wednesday afternoon. There's always quite a good

afternoon. There's always quite a good uh set of people. So obviously Supra is always there. And for Bertton Jones is

always there. And for Bertton Jones is there, Natalie Levina is there. I'm

there. So if somebody's interested in, you know, learning a bit more about GTM, please come along. We're going to have another event in in Liston.

Yeah, I've gone to that a couple times.

SIG SIG is special interest group and we have a few of those and this is the one on grounded theory and yeah, it's the last day of ISIS. I've been there. It's

always well attended.

It's always good.

Yeah, it's a great great uh SIG. It's

one of probably the most persistent one of the most kind of yearover-year ever since I was a PhD student that just exists and it's always well attended right it has a high reputation has great

formats like one of the things you do every year is you take a published GTM paper and deconstruct it with the authors which is incredibly helpful of course for prospective authors so yeah I can I can mirror what Nick said I mean

like from when I was a student this was always one of the most highly reputable workshops in in six that people highly recommend it and that hasn't changed in 20 years. So, thanks for all the work

20 years. So, thanks for all the work you're doing there and thanks obviously also for the work at at JIT. That's a

hell of a job running a journal. You've

been doing it for a long time as well.

So, uh thanks for coming on. I know it's very late for you as well, right? So,

thanks for for definitely out of time also. Yan, you were too modest to

also. Yan, you were too modest to mention that you were selected as best GTM paper I think two years ago and then last year you did the deconstruction uh uh for us which I want to even we had to

do it remotely in that case but I also took quite a uh kind of interesting ideas that you had including putting the data online between the review processes which really helps people following. I

thought that was was an interesting move. I might copy that if I may. Um all

move. I might copy that if I may. Um all

right, thank you so much for having me on.

Thanks for coming on Daniel. Really

appreciate it. Take care.

Thanks, guys. See you.

See you in this one. [music]

See you in this one. [music]